Modeling Surgical Resident Performance

The goals of this work were to evaluate the methods used for assessing surgical residents and to model surgical resident performance. Currently, residents are evaluated by the attending surgeons using a one-page paper evaluation form after each rotation in a particular department. These subjective questionnaires require the evaluators to rate residents in competency areas that are thought to define a successful surgeon. An electronic database was created from resident performance records collected over the past 33 years from the Department of Surgery at the Tufts-NEMC. A usability study examined the effectiveness of the design of the evaluation form and the competency measures. Analysis showed nine changes in format between 1972 and 2005, varying in the competencies rated and rating scales used. Regression analysis was used to model the performance of surgical residents. Results showed that judgment (p < 0.0001), initiative (p < 0.0001), and reaction to stress (p = 0.0206) were significant predictors of a successful outcome. This model may be used to predict the success of new residents and possibly target weaknesses in the surgical education curriculum.